Subject: Re: direct I/O again
To: Chuck Silvers <>
From: Bill Studenmund <>
List: tech-kern
Date: 03/27/2006 19:36:48
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Mon, Mar 27, 2006 at 07:23:34PM -0800, Bill Studenmund wrote:
> On Mon, Mar 27, 2006 at 07:57:50AM -0800, Chuck Silvers wrote:
> > hi folks,
> >=20
> > I've finished the initial implementation of direct I/O,
> > there's an updated diff in:
> >=20
> >
> Could you please move the branch base tag after updates? :-) You seem to=
> have some vmspace changes showing up as diffs, when I expect they are=20
> already in the main tree.
> I'm still looking at the substance of the changes.

Ok, I looked at the substance of it, and it seems fine.

I'm actually quite interested in your mentioning support for cuncurrent=20
writes. What did you have in mind for that?

The two things I see that are needed are:

1) some way to handle keeping the writes out of each others' way. An=20
extent map comes to mind...

2) how do we keep potential allocations out of each others' way? i.e.=20
fine-grained locking on changing the block allocation tables. ??

Taek care,


Content-Type: application/pgp-signature
Content-Disposition: inline

Version: GnuPG v1.2.3 (NetBSD)